Significance analysis of microarrays

SAM: Significance Analysis of Microarrays

Introduction

With the advent of DNA microarrays it is now possible to measure the expression of thousands of genes in a single hybridization experiment. The data generated is considerable and a method for sorting out what is significant and what isn’t is essential. Significance Analysis of Microarrays (SAM) is a statistical technique, establishe in 2001 by Tusher, Tibshirani and Chu, for determining whether changes in gene expression are significantly significant in a set of DNA microarray experiments. SAM identifies statistically significant genes by carrying out gene specific t-tests and computes a statistic d_j for each gene j, which measures the strength of the relationship between gene expression and a response variable ^1,7,8. The response variable describes and groups the data based on experimental conditions. An example of a response variable is an affected group versus a control group for a certain disease with samples from different patients (unpaired grouping) (Chu). In this method, repeated permutations of the data are used to determine if the expression of any gene is significant related to the response. The use of premutation-based analysis accounts for correlations in genes and avoids parametric assumptions about the distribution of individual genes. This is an advantage over other techniques (for example ANOVA and Bonferroni), which assume equal variance and/or independence of genes⁶.

File:SAMkim.png

Running SAM

SAM is available for download online at [1] for academic and and non-academic users after completion of a registration step. System requirements and installation instructions are available on the website following registration.

Data Format ¹

• Data is put into an Excel spreadsheet
  • Column A2 is Titled with the Gene Name
  • Column B2 is Titled with the Gene ID (this is linked to the SOURCE website by SAM-if SOURCE web-site lookup is required a unique identifier (Clone ID, Assession #, or Gene Name/Symbol) ought to be used
  • Remaining Columns contain values with expression measurements for each respective gene
  • Row 1 should contain a coded (numerical values representing response type) value for each expression measurement
    • example 1 = control, 2 = affected

• SAM is run as an Excel Add-In
  • Choose correct values for your data set and click okay
  • SAM Plot Controller allows Customization of the False Discovery Rate and Delta
  • SAM Plot and SAM Output are generated with Significant Genes, Delta Table, and Assesment of Sample Sizes

Permutations and Block Permutations ¹

• Permutations are calculated based on the number of samples
• Block Permutations
  • Blocks are batches of microarrays; for example for eight samples split into two groups (control and affected) there are 4!=24 permutations for each block and the total number of permutations is (24)(24)= 196
  • used for class II unpaired response variables
• a minimum of 1000 permutations are recommended ^{1,2 ,3}
• Amount of permutations is set by the user when imputing correct values for the data set to run SAM

SAM Calculations ^{1,7, 8}

SAM calculates a test statistic for relative difference in gene expression based on permutation analysis of expression data and calculates a false discovery rate. The principle calculations of the program are illustrated below.

File:R and S.jpg

The s_o constant is chosen to minimize the coefficient of variation of d_i

File:FDR.jpg

The SAM algorithm can be stated as ^7,8

1. Order test statistics according to magnitude
2. For each permutation compute the ordered null scores
3. Plot the ordered test statistic against the expected nul scores
4. Call each gene significant if the absolute value of the test statistic for that gene minus the mean test statistic for that gene is greater than a stated threshold
5. Estimate the false discovery rate based on expected versus observed values

SAM Features ^5,6,7,8

• Data from Oligo or cDNA arrays, SNP array, protein arrays,ect can be ultilized in SAM
• Correlates expression data to clinical parameters
• Correlates expression data with time
• Uses data permutation to estimates False Discovery Rate for multiple testing
• Reports local false discovery rate and miss rates
• Can deal with blocked design for when treatments are applied within different batches of arrays
• Can adjust threshold determining number of gene called significant

References

1. Chu, G., Narasimhan, B, Tibshirani, R, Tusher, V. "SAM "Significance Analysis of Microarrays" Users Guide and technical document." [2]
2. Dinu, I. P., JD; Mueller, T; Liu, Q; Adewale, AJ; Jhangri, GS; Einecke, G; Famulski, KS; Halloran, P; Yasui, Y. (2007). "Improving gene set analysis of microarray data by SAM-GS." BMC Bioinformatics 8: 242.
3. Jeffery, I. H., DG; Culhane, AC. (2006). "Comparison and evaluation of methods for generating differentially expressed gene lists from microarray data." BMC Bioinformatics 7: 359.
4. Kooperberg, C., S. Sipione, et al. (2002). "Evaluating test statistics to select interesting genes in microarray experiments." Hum. Mol. Genet. 11(19): 2223-2232.
5. Larsson, O. W., C; Timmons, JA. (2005). "Considerations when using the significance analysis of microarrays (SAM) algorithm." BMC Bioinformatics 6: 129.
6. Tusher, V. G., R. Tibshirani, et al. (2001). "Significance analysis of microarrays applied to the ionizing radiation response." Proceedings of the National Academy of Sciences 98(9): 5116-5121.
7. Zang, S., R. Guo, et al. (2007). "Integration of statistical inference methods and a novel control measure to improve sensitivity and specificity of data analysis in expression profiling studies." Journal of Biomedical Informatics 40(5): 552-560
8. Zhang, S. (2007). "A comprehensive evaluation of SAM, the SAM R-package and a simple modification to improve its performance." BMC Bioinformatics 8: 230.